VR applications for medical simulations such as emergency situations offer new ways of distributing knowledge and providing practical skills for saving a life. Compared to the genre of serious games in which players are trained as exports or layperson first-aiders within an interactive gameplay, VR applications potentially represent an even more complex communicative environment. Additional semiotic resources such as touch, body movements and proprioception (cf. Martin et al. 2022) are used to construct the virtual world and instruct players to perform certain procedures. From a multimodal perspective (cf. Bateman et al. 2017), the communicative situations constructed by the virtual environment bring with them an increasing level of interactivity and ergodicity and it is particularly challenging to address these analytically. Especially the cyberphysical infrastructure provided in these applications add to the complexity by embedding holograms, 360-degree videos, etc. which need to be taken into consideration in the analysis of the instructional techniques. While there is now research on e-learning strategies such as serious games (e.g. Boada et al. 2020) and mobile apps (e.g. Metelmann et al. 2018), not much work has been done with regard to the (multimodal) design of VR applications and how these can be used to teach resuscitation and other procedures. It will therefore be interesting to address this question from a qualitative perspective with future potential for empirical research on the effectiveness of these applications. In this talk, we will address the analytical challenges by looking at two different examples of VR applications for first aid (Lifesaver VR, Resuscitation Council UK; Basic Life Support, Dual Good Health) and providing a foundational framework for the multimodal analysis of the communicative situations created in these two applications based on our previous work on video games (Wildfeuer & Stamenković 2022). The data to work with are recordings of the initial usage of the applications as provided by the production companies. We aim to not only show how recent developments in multimodality research are well-equipped for the effective analysis of these artifacts, but also how the analytical results can be practically implemented in the further development of these applications.